Martin Krzywinski / Genome Sciences Center / Martin Krzywinski / Genome Sciences Center / - contact me Martin Krzywinski / Genome Sciences Center / on Twitter Martin Krzywinski / Genome Sciences Center / - Lumondo Photography Martin Krzywinski / Genome Sciences Center / - Pi Art Martin Krzywinski / Genome Sciences Center / - Hilbertonians - Creatures on the Hilbert Curve
Poetry is just the evidence of life. If your life is burning well, poetry is just the ashLeonard Cohenburn somethingmore quotes

In Silico Flurries: Computing a world of snow. Scientific American. 23 December 2017

art + science activism

Watch the video of this project, which features the participants who have a BRCA mutation and their interaction with the piece. The video also highlights the design and construction of the mural.

Martin Krzywinski @MKrzywinski
Science and art and personal stories of cancer survivors combine into this beautiful depiction of the complexity and individuality of the genome. (Free the Data)

Human Genome Art by Humans with Genomes

I recently took part in a deeply meaningful collaboration of science, art and personal stories of cancer survivors.

Together with Joanna Rudnick and Aaron De La Cruz, we sought to create a work of art that combines the science of cancer genomics and the individuals whose lives are affected by genetic mutations in the BRCA1 and BRCA2 genes, where genomic changes drastically increase one's chances of breast and ovarian cancer.

We wanted to make something that is scientifically accurate, artistically beautiful and emotionally engaging. The complexity of the genome, the multitudes of other genes and possible mutations and the millions of personal stories of hardship and survival were just a few of the elements we wanted to include the the piece.

My role was to provide the scientific direction behind the design and incorporate it into the aesthetic of Aaron De La Cruz, a street artist from San Francisco whose work echoes information, complexity, interaction and continuity. We all have a genome — a different genome. The ways in which our genomes are different is what gives us traits like hair and eye color, but is also what makes some of us predisposed to diseases like cancer.

The mural, which includes elements drawn by the cancer survivors, is part of the Free the Data campaign, which is advocating for an open access model of genome mutation databases so that scientists everywhere can analyze it and help women make informed choices about their breast-cancer risk.

The piece Importance of Data Sharing by Nature Methods illustrated the point:

Imagine you are a physician or researcher and seek to get more confirmation on the clinical impact of particular genetic variants. If your search of public databases comes up empty this does not necessarily mean that nothing is known about the mutations in question. Rather, the information may be locked away as a trade secret in a genetic testing company’s proprietary database.

The New York Times article DNA Project Aims to Make Public a Company’s Data on Cancer Genes captures the current state of the situation.

The mural was constructed on location at InVitae in San Francisco.

A video of the project is available.

Beautiful, meaningful and personal

This work will be, as far as I know, the first human annotation of mutations in the human genome by humans whose genomes have the mutations. That's quite a term!

I've always been mindful of the necessity of the mingling of art and science. In my work I tried to add things I felt about the science I thought to create work that combines our objective understanding of the world we live in with the subjective experience of living in it. This project, by far, has been the most keenly felt.

Martin Krzywinski @MKrzywinski
Adding emotion, keeping the science. (Free the Data)

the design

The mural was created in San Francisco on Saturday, July 13th, 2013. We are starting with a 11' x 6' wood canvas. These dimensions reflect the ratio of lengths of BRCA1 and BRCA2 proteins (1,863 and 3,418 amino acids, respectively)

Martin Krzywinski @MKrzywinski
The canvas aspect ratio reflects the ratio of BRCA1 and BRCA2 protein lengths. The proteins are represented on the canvas as lines. (Free the Data)

The BRCA1 and BRCA2 proteins are drawn on the canvas as straight-line sections.

Martin Krzywinski @MKrzywinski
The genes are depicted on the canvas as their protein products. (Free the Data)

The locations of the participants mutations are positioned on the protein lines as circles. For individuals with large deletions, the circle is placed at the first affected amino acid. Because BRCA1 is location on the opposite strand (anti-sense), its start on the canvas is on the right.

Martin Krzywinski @MKrzywinski
11 mutations, one for each of the cancer previvor and survivor participants, are placed on the protein lines as circles. The start of BRCA1 is on the right to reflect that this gene is on the anti-sense strand. (Free the Data)

The rest of the genome is now drawn. Aaron's style is perfect for depicting information and the endless complexity of the genome and its interacting elements. We were careful to include elements that indicate that the story told today is not complete. Millions of others have mutations in thousands of other genes, each potentially life-threatening. Just as the stories of our participants will continue to evolve, other stories are waiting to be told.

Martin Krzywinski @MKrzywinski
BRCA1 and BRCA2 proteins and their mutations, together with the rest of the genome. Other lines and circles hint at other genes, other mutations, as well as the biochemical interactions in the cells and personal interactions of those affected by the mutations. (Free the Data)

Once the "reference" genome is depicted, participants with BRCA1 and BRCA2 mutations will complete the art work by individually marking the positions of their mutations on the art using personalized colors. With Aaron's help, everyone created their own color by mixing primary colors.

Martin Krzywinski @MKrzywinski
Participants fill in their mutation circles with their personalized color. (Free the Data)

From base pair, to genome, to person, to life. All it takes is one tiny change in the genome to change a life forever.

Martin Krzywinski @MKrzywinski
The mutations of 11 people in the vastness of the genome. What's your story? (Free the Data)

creation of free the data mural

The BRCA1 and BRCA2 lines were placed on the canvas by first pinning two pieces of string, marked with the positions of the mutations.

Martin Krzywinski @MKrzywinski
String was used to mark the placing of lines and mutations. (Free the Data)

After drawing the protein lines, it was time to fill the canvas.

Martin Krzywinski @MKrzywinski
Aaron De La Cruz creating the art work. Here, he is filling the space in the canvas around the BRCA1 and BRCA2 segments with his design. The project was shot with a Red Camera—this is a sequence from its render application. (Free the Data)

Over the next 4 hours, Aaron filled in the canvas with the "rest" of the genome.

Martin Krzywinski @MKrzywinski
Aaron De La Cruz creating the art work. Here, he is filling the space in the canvas around the BRCA1 and BRCA2 segments with his design. The project was shot with a Red Camera—this is a sequence from its render application. (Free the Data)


Lucy, Karen, Steve, Ghecemy, Joanna, Jill, Lisa, Lynn, Ruth, Jenica, Susan

Cancer previvors and survivors who have been diagnosed with a mutation on BRCA1 or BRCA2 genes.

Joanna Rudnick (director/producer)

Joanna made her directorial debut with the Emmy-nominated In the Family, a deeply personal film about coming to terms with testing positive for the breast cancer gene BRCA1 mutation and following the storylines of other women and families facing the same hard choices. In the Family premiered at Silverdocs in 2008, was broadcast nationally on PBS P.O.V. the same year and was a finalist for the NIHCM Foundation’s Health Care Radio and Television Journalism Award.

Joanna received a master’s degree in Science and Environmental Journalism from New York University and a bachelor’s degree in English from Northwestern University. Joanna loves the opportunity to teach and mentor and served as an adjunct professor at Northwestern University’s Medill School of Journalism in the past.

She has written for several publications including Audubon Magazine, The Artful Mind, The Berkshire Record and Humanities. Before finding her way to the wonderful world of documentaries, Joanna served as an Americorps volunteer, implementing project-based environmental curricula in the San Francisco Public School System.

Joanna is one of the cancer survivors whose mutations are encoded in the art.

Aaron De La Cruz (artist)

Aaron De La Cruz's work, though minimal and direct at first, tends to overcome barriers of separation and freely steps in and out of the realms of design, graffiti, and illustration.

The parameters he has chosen to work within actually allow him to free himself and react to the very limitations he has created. This overriding structure and the lack of deliberation while moving within creates a tension when encountering his work due to the almost computer generated grid like systems he creates by unplanned markmaking. The act and the marks themselves are very primal in nature but tend to take on distinct and sometimes higher meanings in the broad range of mediums and contexts they appear in and on.

Martin Krzywinski @MKrzywinski
Work by Aaron De La Cruz. (Aaron De La Cruz)

His work finds strengths in the reduction of his interests in life to minimal information. De La Cruz gains from the idea of exclusion, just because you don't literally see it doesn't mean that its not there.


news + thoughts

Curse(s) of dimensionality

Tue 05-06-2018
There is such a thing as too much of a good thing.

We discuss the many ways in which analysis can be confounded when data has a large number of dimensions (variables). Collectively, these are called the "curses of dimensionality".

Martin Krzywinski @MKrzywinski
Nature Methods Points of Significance column: Curse(s) of dimensionality. (read)

Some of these are unintuitive, such as the fact that the volume of the hypersphere increases and then shrinks beyond about 7 dimensions, while the volume of the hypercube always increases. This means that high-dimensional space is "mostly corners" and the distance between points increases greatly with dimension. This has consequences on correlation and classification.

Altman, N. & Krzywinski, M. (2018) Points of significance: Curse(s) of dimensionality Nature Methods 15:399–400.

Statistics vs Machine Learning

Tue 03-04-2018
We conclude our series on Machine Learning with a comparison of two approaches: classical statistical inference and machine learning. The boundary between them is subject to debate, but important generalizations can be made.

Inference creates a mathematical model of the datageneration process to formalize understanding or test a hypothesis about how the system behaves. Prediction aims at forecasting unobserved outcomes or future behavior. Typically we want to do both and know how biological processes work and what will happen next. Inference and ML are complementary in pointing us to biologically meaningful conclusions.

Martin Krzywinski @MKrzywinski
Nature Methods Points of Significance column: Statistics vs machine learning. (read)

Statistics asks us to choose a model that incorporates our knowledge of the system, and ML requires us to choose a predictive algorithm by relying on its empirical capabilities. Justification for an inference model typically rests on whether we feel it adequately captures the essence of the system. The choice of pattern-learning algorithms often depends on measures of past performance in similar scenarios.

Bzdok, D., Krzywinski, M. & Altman, N. (2018) Points of Significance: Statistics vs machine learning. Nature Methods 15:233–234.

Background reading

Bzdok, D., Krzywinski, M. & Altman, N. (2017) Points of Significance: Machine learning: a primer. Nature Methods 14:1119–1120.

Bzdok, D., Krzywinski, M. & Altman, N. (2017) Points of Significance: Machine learning: supervised methods. Nature Methods 15:5–6.

...more about the Points of Significance column

Happy 2018 `\pi` Day—Boonies, burbs and boutiques of `\pi`

Wed 14-03-2018

Celebrate `\pi` Day (March 14th) and go to brand new places. Together with Jake Lever, this year we shrink the world and play with road maps.

Streets are seamlessly streets from across the world. Finally, a halva shop on the same block!

Martin Krzywinski @MKrzywinski
A great 10 km run loop between Istanbul, Copenhagen, San Francisco and Dublin. Stop off for halva, smørrebrød, espresso and a Guinness on the way. (details)

Intriguing and personal patterns of urban development for each city appear in the Boonies, Burbs and Boutiques series.

Martin Krzywinski @MKrzywinski
In the Boonies, Burbs and Boutiques of `\pi` we draw progressively denser patches using the digit sequence 159 to inform density. (details)

No color—just lines. Lines from Marrakesh, Prague, Istanbul, Nice and other destinations for the mind and the heart.

Martin Krzywinski @MKrzywinski
Roads from cities rearranged according to the digits of `\pi`. (details)

The art is featured in the Pi City on the Scientific American SA Visual blog.

Check out art from previous years: 2013 `\pi` Day and 2014 `\pi` Day, 2015 `\pi` Day, 2016 `\pi` Day and 2017 `\pi` Day.

Machine learning: supervised methods (SVM & kNN)

Thu 18-01-2018
Supervised learning algorithms extract general principles from observed examples guided by a specific prediction objective.

We examine two very common supervised machine learning methods: linear support vector machines (SVM) and k-nearest neighbors (kNN).

SVM is often less computationally demanding than kNN and is easier to interpret, but it can identify only a limited set of patterns. On the other hand, kNN can find very complex patterns, but its output is more challenging to interpret.

Martin Krzywinski @MKrzywinski
Nature Methods Points of Significance column: Machine learning: supervised methods (SVM & kNN). (read)

We illustrate SVM using a data set in which points fall into two categories, which are separated in SVM by a straight line "margin". SVM can be tuned using a parameter that influences the width and location of the margin, permitting points to fall within the margin or on the wrong side of the margin. We then show how kNN relaxes explicit boundary definitions, such as the straight line in SVM, and how kNN too can be tuned to create more robust classification.

Bzdok, D., Krzywinski, M. & Altman, N. (2018) Points of Significance: Machine learning: a primer. Nature Methods 15:5–6.

Background reading

Bzdok, D., Krzywinski, M. & Altman, N. (2017) Points of Significance: Machine learning: a primer. Nature Methods 14:1119–1120.

...more about the Points of Significance column

Human Versus Machine

Tue 16-01-2018
Balancing subjective design with objective optimization.

In a Nature graphics blog article, I present my process behind designing the stark black-and-white Nature 10 cover.

Nature 10, 18 December 2017